April 20, 2015

Australian Markets Weekly: Can we believe the ABS labour market data?

That is a key question for investors seeking to work out whether the RBA will ease again at the May Board meeting and furthermore whether the market is correct in pricing nearly two full interest rate cuts by February 2016

  • Key events this week are a speech by the RBA Governor in NY tonight (230am Tuesday Sydney time), the RBA Minutes Tuesday at 11.30am and the Q1 CPI on Wednesday at 11.30am
  • Last week provided two upside data surprises – the NAB survey and Australian labour force data. Both suggest a reduced chance of a further rate cut at the May Board meeting given the non-mining economy is beginning to lift
  • A key question, which we investigate in detail in this report, is how believable are the ABS labour market data? Our investigation suggests employment has strengthened (since late last year) and may now be growing at a pace fast enough to stabilise unemployment around the 6.25% level, if current trends continue. If realised, this would be quite a different picture to the RBA’s forecasts of February and to the RBA, NAB and the markets’ forecast that unemployment will continue to rise toward the 6.5-6.6% region.
  • Our key market themes remain for a moderately stronger US$, moderately weaker $A and moderately higher term yields by year end. The possibility of no further cut in rates by the RBA at the May meeting, may see the $A a little higher in the short term, but medium term, a higher US$ is still expected to deliver a $A around US$0.73/0.74 toward year end. Term yields are expected to rise as the US raises interest rates, however, very low (and in some cases negative) yields in Europe are likely to broadly constrain the rise in yields.

Can we believe the ABS labour market data?

That is a key question for investors seeking to work out whether the RBA will ease again at the May Board meeting and furthermore whether the market is correct in pricing nearly two full interest rate cuts by February 2016. We’ll go into the answer in more detail below, however, the short answer is that our analysis of the raw employment data suggests a strengthening in employment appears to have occurred over the past 4-5 months and while the February data is probably a little overstated, the strength of the March read tends to confirm the strengthening in employment growth now suggested by the ABS data.

This seems quite significant, since instead of the economy only generating sufficient jobs to keep unemployment on a slowly rising trend, in fact the economy could already be generating sufficient jobs to actually stabilise and even marginally reduce unemployment. This is quite a different situation to that which the RBA was considering before it eased interest rates in February – the full effects of that easing are still to flow through to the economy. As such, the latest employment data at face value present a significant challenge to the RBA’s assessment revealed in the March Board Minutes, when the Board was waiting for more information to confirm whether the economy was indeed on the previously forecast (sub-trend) growth track – more on that analysis below. That said, the Bank will consult a broad range of labour market indicators, most of which suggest a more moderate improvement in recent months.

The Week Ahead

This week’s action in Australia revolves first around a speech by RBA Governor Stevens in NY (at 2.30am Tuesday morning Australian time) – no title available at this time, the Minutes to the April RBA Board meeting, released on Tuesday at 11.30am, followed by the Q1 CPI on Wednesday. The quarterly NAB Business survey will also provide a more detailed update on business capital expenditure intentions, for which there have been a few encouraging signs in recent times. The RBA mentioned a pick-up in business lending in its April Interest Rate Decision.

The Minutes may provide further clues as to whether the RBA will ease further at its May Board meeting, which remains the near universal expectation of economists, but which now is only around 55% priced by markets. As our analysis below suggests, we are less confident than before the stronger NAB business survey of early last week and the stronger employment data of recent months.

Labour force data believable

The big surprise of last week was considerably stronger than expected Australian labour market data for March. This was actually the second significant upside data surprise of the week, with the March NAB Business Survey also showing an improvement in business conditions, particularly in the large (non-mining) states of NSW and Victoria, both of which are now recording quite reasonable business conditions. The labour force release then revealed that 37,700 jobs were created in March, following an upwardly revised 42,000 jobs gain in February (initially +15,600). Together, these gains were sufficient, even with a 0.2 percentage point jump in the participation rate, to bring the unemployment rate down from 6.3% in January to 6.1% in March. Employment is now growing in trend terms at nearly 21,000 a month, sufficient to generate a small fall in the number unemployed of 1,500 persons per month and to stabilise the unemployment rate at 6.1%. (For more information refer to Chart 1 of the pdf).

The reported strengthening in employment has been occurring since around October/November last year, in spite of the negative developments on the commodity price front and concerns with the mining sector and WA economy in particular. Interestingly, at this stage WA continues to record slightly positive trend employment growth of around 2,500 persons per month, though job ads have weakened further in recent months, suggesting employment growth will likely weaken from here. It is also worth noting that this broad strengthening in national employment growth was occurring before the RBA’s interest rate cut in February. The current rate of growth of employment would seem at face value to contradict the RBA’s February forecast of a further period of sub-trend economic growth. This forecast of course predicated the February rate cut.

A key issue, therefore, is how believable is the March (and other recent months’) labour market data? Assessment of this question is complicated by the general lack of reliability of the Labour Force Survey in recent months, and specifically, by the comment in the March report that:

“The February 2015 seasonally adjusted estimates include an adjustment for changes to the timing and content of the supplementary survey program; this adjustment has been revised following the availability of March 2015 data. The independent technical review of the Labour Force Survey noted that there was the possibility of instability in seasonally adjusted estimates in months with changes to the supplementary survey program, such as February.”

NAB contacted the ABS to discuss what this comment actually meant and our interpretation follows. In essence, once the March data again printed a very strong outcome, the ABS, decided that more of the very large increase in original February employment was real and not simply due to the presence of the supplementary survey program, which boosted “employed” responses significantly in August 2014. Previously the ABS had made a prior adjustment to effectively suppress (remove) much of the strength in the February original data.

Forgive, the depth of examination, however, given the validity or otherwise of the current signal of the employment data could mean that the RBA will not cut rates again, we feel it is worth the excursion! The original (non-seasonally-adjusted, no prior adjustment) data for employment is plotted for successive years. The broad seasonal pattern for employment is very obvious, in particular the normal drop in employment levels in August. The presence of a supplementary survey in August 2014 is thought to explain the very different seasonality from the norm that occurred last year and produced that memorable August labour force gain of around +121,000.

The ABS of course picked this up when the compilation of the September data suggested an even larger fall would result – and then temporarily suspended publication of the seasonally-adjusted data. February 2015 also contained a supplementary survey (and also produced a much higher level of original employment than would normally be expected). The ABS effectively removed much of this effect in its prior seasonal adjustment for known factors (there are three types of these prior adjustments: (i) for the variable timing of Easter; (ii) for the variable timing of the start of the labour force survey – eg in January this starts later than normal due to holidays; and now (iii) for the presence of supplementary surveys). However, when the March data effectively confirmed the strength of the February original reading, the ABS reduced the estimated correction (suppression) for the supplementary survey effect and hence the February data was revised up sharply. The ABS now believes more of that strength was real – and it is concentrated in the large non-mining states of NSW and Victoria. (For more information refer to Chart 2 of the pdf).

The April data will of course provide a further guide as to how robust this signal is. However, another way to analyse the trend, is to consider the spacing between the lines in successive years. When the lines are close together, there is slow employment growth y/y. A wider gap between successive years indicates strong employment growth. This reported strengthening in employment growth has been occurring since around the October/November timeframe. As such it seems a strong signal to us, and one that is at face value, inconsistent with the RBA’s February assumption of sub-trend economic growth and a continuing gradual rise in the unemployment rate. This assumption precipitated the February rate cut and the March Minutes guided that the RBA was awaiting further data to confirm that the economy was indeed on this sub-trend growth path before easing further. To us, this raises a significant risk that the RBA will again disappoint the market in May, especially given the Bank has already eased further in February. It particularly makes unlikely the prospect of there being multiple further easing moves in the next 3-6 months (the market has been reducing this pricing in recent sessions).

Of course given the issues the ABS has had with its labour market data, the RBA would consult the message of a wide range of labour market data. There were two such updates released in the past week. The NAB Business Survey produced a zero net balance for employment, admittedly not a strong reading, but one which in conjunction with trends in job advertising, suggests the economy is currently creating around 15,000 jobs per month. This estimate suggests the ABS survey is currently slightly overstating the pace of jobs growth, however, growth in employment has still strengthened moderately on this measure. (For more information refer to Chart 3 of the pdf).

SEEK also released its monthly job ads data last week. Job ads rose 0.2% in March and continue to record moderate growth (+6.7% y/y). The state breakdown of the SEEK data shows the continuing strengthening in job advertising in NSW and Victoria (confirmed in both the ABS employment data and the NAB business survey) and the recent renewed decline in WA job advertising (-12.4% y/y). (For more information refer to Chart 4 and 5 of the pdf).

This week

This week is all about the April Board Minutes and the Q1 CPI. The April Minutes may give some further clues as to the RBA’s inaction in April, which may in turn allow more confident assessment of the chances of a further rate cut at the May Board meeting. The May Board meeting has been our preferred meeting for contemplation of a further easing move, by virtue of it being a full forecasting round, while also allowing sufficient time for the Board to both assess the impact of the recent easing and the general trend in the economy. On the basis of our employment analysis, plus the continuing strength in housing markets (another record clearance rate in Sydney on the weekend), support for a further easing is reduced.

The markets are likely to see another very low CPI as supporting the case for further easing. NAB has a headline forecast close to the market expectation of a flat outcome, in large part due to seasonal effects and the impact of large petrol price falls. Our models are generating a low forecast 0.3/0.4% q/q underlying rate, which suggests downside risk to the markets’ forecast of a 0.5/0.6% q/q underlying rate. NZ’s low -0.3% q/q headline Q1 CPI outcome supports the expectation of a low headline CPI. However, as inflation is comfortably at the low end of the RBA’s target band, we do not see inflation currently as a binding constraint. The Bank’s view on activity and its forecasts are much more important at this stage.

Before both of these events, RBA Governor Stevens speaks in NY at 2.30am Australian time on Tuesday morning.

Market themes

NAB’s core themes on markets have been:

  • The US$ is expected to continue to benefit moderately from the stronger US economy and expectations of the Fed beginning to raise interest rates in the June to September timeframe;
  • The $A/US$ exchange should remain under further moderate downward pressure as the US$ increases, though much of the adjustment may have already occurred in the near term; and
  • The RBA is expected to deliver a further interest rate cut in the next month or two, but after that, keep interest rates on hold for an extended period. Term yields are expected to be under moderate upward pressure over the medium term from higher US yields.

Recent market and economic developments provide for some moderate near-term challenges and variations around those themes, though not necessarily long-term challenges. The Australian data prints have generally suggested the non-mining economy is beginning to strengthen, which reduces the chances and potentially the timing of a further interest rate cut. Because bond and FX markets continue to price with nearly two interest rate cuts over the next year, this raises the risk that the $A may move a little higher in the near term, though the expectation remains medium term that a stronger US economy and US$ will produce a lower $A by year end (to around US$0.73/0.74). And while the US Fed is likely to tighten interest rates twice before year end, the continuing rally in European long yields is offsetting much of the expected negative impact on term yields in the US and Australia in the near term. By year end, however, we still expect to see moderately higher US and Australian term yields.

For full analysis, download report:

For further FX, Interest rate and Commodities information visit nab.com.au/nabfinancialmarkets